Reprinted from the U. S. Society’s The Keeper’s Log – Winter, 2016 Weights and Weightways at American By John T. Graham

Crooked River Lighthouse. USLHS archive photo. It had been a long January 1927 night, and Keeper Thorwald Hansen at the Crooked River, Florida, Lighthouse was tired. His assistant was ill and so the nightly watch fell to him alone. The fourth-order Fresnel was rotated every 12.5 seconds by a clockwork driven by the 150 pounds of iron weights that hung on the end of a cable. Keeper Hansen had let these weights descend almost to the bottom of the tower, which took about 12 hours, so now he would have to wind them 80 feet back to the top before the lens came to a halt.

34 Clockworks—Winter 2016 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 he bright light that beamed from the “winding of the weights,” just as it put an the tower of a lighthouse, whether end to the nightly lighting of the lamp. it was a tall one near the shore or a Today just a handful of lighthouses still short one high on a headland, was have a Fresnel lens that rotates by use of a a welcoming, comforting, and nec- weighted clockwork; many others still display essary beacon to the mariner navigating the remnants of such a system, some with the often treacherous waters of the coastal United weights rusting away in the weightway; but in States. The single most important service that the vast majority of cases, the weights them- a lighthouse provided to that mariner was to selves disappeared long ago when the electric confirm that “You are here” along those oth- motor relegated them to the junk pile. The erwise dark and foreboding shores. following lighthouses still have an operational It was not enough for that light to be only lens that is rotated by a weight-driven clock- bright, as the ship master would then still be work: Split Rock, Minnesota; , wondering which lighthouse he was viewing. ; Elbow Reef, Abaco, Bahamas (the This was especially true where the lights from Elbow Reef light still uses a kerosene lamp, several different lighthouses could be seen from perhaps the last in the world to do so); Dixon a given location. Thus, from the earliest dates, Hill, San Salvador, Bahamas; La Martre, Que- lighthouse administrators sought to make the bec. There may, of course, be others. light from each tower somehow unique. Each In the typical system, it was a “rope” light was to have its own “signature” or “char- (in earlier days made of fiber, in later days acteristic” that set it apart from its neighbors. Weight at Pigeon Point, California. Photo by made of iron/steel) that was wound onto the With the advent of the use of the highly ef- Thomas Tag. drum of the clockwork. This rope held the ficient Fresnel in the lantern rooms, one weights. The amount of weight needed was of the most common methods of achieving In the days before electricity and electric mo- determined by the weight of the lens to be this was to make some lenses appear to flash. tors came to the shores, it was a weight-driven turned, the speed at which it needed to ro- The technical differences between lenses gearbox that provided the motive power to turn tate to produce the specific characteristic as- that exhibited a flashing characteristic and even immensely heavy Fresnel lenses to produce signed to that light, and the gearing ratios of those that did not are explained well by many those flashes. For a technical look at these sys- the clockwork itself. This all went into deter- others in many other sources. For my pur- tems, refer to “The Clock Without Hands” by mining how fast the weights would descend. poses here, suffice it to say that most lenses Thomas Tag, which can be found in the Spring Lighthouse engineers then had to determine that exhibited a light that appeared to flash 2008 issue of this journal. Another great source how far down the tower they could or should had to turn, or rotate. It was the passing of is the U.S. Lighthouse Society’s website, specifi- descend and at what point the keeper on each of the “bull’s eye” sections of the lens cally www.uslhs.org/clockworks. Information duty needed to wind them up again. between the lamp in the epicenter and the on the various types of clockworks and their Although this article focuses on the eye of the viewer, however far away he might operation can be found in these and many oth- weight systems used in the Fresnel lens era, be, that created the appearance of the flash. er sources. This article will focus on the weights lamp assemblies used before the advent of that drove the gears those revolutionary lenses often were rotat- and the paths they ed as well, using a similar, if simpler, system traveled from top to to attempt to produce a distinctive charac- bottom of the light teristic for particular lights. towers. The earlier lighthouses utilized wick lamps Keeper Han- with a reflector to direct more of the light sen, noted in the forward. There were often several to many introductory para- of these lamps in a given lighthouse, usually graph, was simply mounted in a “chandelier” arrangement. Ma- performing one of jor seacoast lights might have had 20 or more the many normal individual lamps in the arrangement, which tasks of a light- were arranged so that the chandelier was two- keeper, tasks that sided, three-sided, or even four-sided. To the had not changed mariner, these lights would show an “occult- for decades. The ing” pattern of light followed by lesser light arrival of electricity and/or darkness, the interval being deter- to the remote sites mined by the arrangement of the lamps. where most light- Little is known about the weights or Weights visible in drop tube window, originally an access door. Split houses are located weightways in the early towers, but as the Ar- Rock, Minnesota. Photo by the author. brought an end to gand-type lamps used in those days weighed

The Keeper’s Log—Winter 2016 35 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016

Much of the informa- Any blank space indicates that the inspect- tion for this article was ob- ing officer did not complete that section of tained directly from staff the report. For my purposes here, I was most at the various lighthouses interested in the following categories: how noted. Unfortunately, much weight was used and how frequently since it has been decades the weights were wound. since a clockwork system How Much Weight Was was used at most light- houses, information about Used? them is simply unknown t was my naïve expectation that larger or not remembered. Ilenses would have required more weight, The other sources of but this was clearly not the case. Seventy-five information used are of the inspection reports reviewed noted the actual lighthouse inspec- amount of weight in use. These ranged from tion reports from the a low of 30 pounds for a third-order lens National Archives, as at Ameila Island, Florida, to a high of 503 Centered drop tube, Split Rock, Minnesota. Photo by Lee Radzak. compiled into a database pounds for a second-order lens at Kilauea by volunteers of the U.S. Point, Hawaii. Twenty-two of the reports relatively little compared to even the smaller Lighthouse Society. Although hundreds noted weights of 200 pounds or more. order Fresnel lenses, correspondingly less have thus far been transcribed, hundreds Although the radius of any given Fresnel weight would have been required. When the more remain to be so transcribed. Although lens is the measure of its “order,” other dimen- first use of a clockwork to rotate a chandelier of there is great inconsistency in the work of sions, and thus the total weight, can vary sig- lamps was put into use will be left to other his- the various inspectors over the years, these nificantly in lenses of the same order. Another torians. Most likely the weights simply hung reports comprise an invaluable glimpse factor is the rotational system in use. Some down from the clockwork, but a reference to into lighthouse operation. [USLHS.org; lenses used a chariot wheel system; some ro- an actual weightway is found in the archives Resources; Lighthouse Databases; then En- tated on a ball-bearing plate; still others turned of the Cape San Blas, Florida, Lighthouse. ter the National Archives Database.] Data on a bed of mercury. There are likely other In an 1848 letter to the district supervisor at from these reports is shown on page 38. variables of which I am not aware. Apalachicola, Light-House Establishment Su- The following is a summary of the in- perintendent Stephen Pleasonton requested dividual weight amounts noted for various that the plans for the staircase and lantern deck orders of lenses: for the lighthouse then under design for this Hyper-radial lens: location be modified to allow for “clockwork Makapuu, Hawaii, 121 pounds and weights.” Such systems had undoubtedly First-order lens: already been in use for decades if not longer. low at Pensacola, Florida, 80 pounds; What did the weights look like? There is high at Cape Fear North Carolina, no single answer to this question; as in al- 500 pounds most all aspects of lighthouse design, there is Second-order lens: a tremendous variety. However, the weights low at Minot’s Ledge, Massachusetts, used with the Fresnel lenses can be generally 100 pounds; high at Kilauea, Hawaii, grouped into three types. 503 pounds Many of them were simply a single, solid Third-order lens: piece of metal, or in some cases of stone, low at Amelia Island, Florida, 30 concrete, or other material, constructed pounds; high at Cape Spencer, Alas- around a metal rod with an eye for connect- ka, 270 pounds ing to the supporting cable. Three-and-a-half-order lens: Many others were a similar solid piece but low at Chandeleur, Louisiana, 60 then had the addition of several smaller indi- pounds; high at Brazos River, Texas, vidual pieces. The ability to add or take away 165 pounds these additional, usually disc-shaped pieces, al- Fourth-order lens: lowed for adjustment to the speed of descent. low at Hendricks Head, Maine, 40 This would have been particularly necessary if pounds; high at Barbers Point, Hawaii, the characteristic of a lighthouse was modified 278 pounds or if a new type/size lens was installed. Fifth-order lens: Still others are made up entirely of stack- Photo shows both the upper access door and low at Butler Flats, Massachusetts, able, circular iron discs of various thickness the weight-resting shelf below. White River, 100 pounds; high at Port Ferro, Puer- and thus weight. Michigan. Photo by Matt Varnum, SPLKA. to Rico, 200 pounds 36 Clockworks—Winter 2016 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 Reports from the earlier years of the Fresnel a reputation for irregularity in the past… era seldom noted any data about weights, rota- The first night I was there it ran 15 sec- tion times, or drop tubes. In a few later instanc- onds fast and required a reduction in es, reports noting weight data from the same driving weight. The next night it ran as much as 25 seconds slow and required lighthouse were found for more than one year. additional weight. The third evening it In some of these the information was identi- stopped twice on us without apparent cal in all years noted, but in others changes in cause and it required two hours close the amount of weight used were documented attention before it braced up. After that from one point in time to another. it ran all right, all night. The Barbers Point, Hawaii, Lighthouse Such irregularity was undoubtedly an is such an instance. Reports for 1912, 1916, occasional situation at all sites with a clock- and 1927 show weights of 182, 278, and 110 work and was one more reason why diligent pounds respectively, all rotating a fourth-or- watchfulness was required of the keepers der lens every five seconds. A history of this The time of lens rotation was critical to lighthouse notes that the fourth-order lens producing the characteristic flash published was installed in 1912 with a five-second rota- in mariners’ light lists, so much so that keep- tion, but notes no changes until a new tower ers were issued precision watches by which to was constructed in 1933, so I can not explain monitor these times. Even a second of variation the need for different amounts of weight. could confuse a ship captain as to his location. Two of the reports are for the Molokai, The range of times it took for a given lens Hawaii, Lighthouse. In 1910, 272 pounds of to make one complete turn is another in- weight turned the lens once every 40 seconds; teresting category of information found in in 1927 only 166 pounds of weight are noted, the reports. Rotation times varied from five yet the lens turned once every 20 seconds. This seconds for a fourth-order lens at Barber’s author can only speculate about these seem- Point, Hawaii, to nine minutes at both St. ingly drastic changes in the amount of weights Augustine, Florida, and Cape San Juan, in use and their effect on lens operation. Puerto Rico. Two hundred and fifty pounds It is important to note that the reports of weight turned a first-order lens every 18 reviewed represent only a fraction of those seconds at the Graves, Massachusetts, Light- that are available in the archives and that house. The same amount of weight turned only a relatively few of those that were re- Plan of original rod system, Cape Hatteras a second-order lens every six minutes at the viewed noted the actual amount of weight North Carolina. Courtesy of John Havel. Petit Manan, Maine, Lighthouse. in use, so the ranges of weights referenced There must be a direct correlation between here may well have been exceeded at both perature and humidity changes could require the size/weight of lens, the mechanical system the low and high ends at other locations. that the angles of these vanes be modified as in use, and the desired rotational time that In one instance obtained from other than needed. This was a job that only experienced went into the determination of how much the inspection reports, current staff at Elbow and practiced keepers of a specific lens/clock- weight was used, but I will leave that explana- Reef, Bahamas, which still turns its first-or- work could accomplish satisfactorily. tion to others. der lens with a clockwork, In spite of the best engineering and top 700 pounds of weight are reportedly in use. quality equipment, sometimes the keepers This is by far the largest amount of weight just couldn’t get things encountered anywhere. to work correctly. In a How much weight was needed to turn a 1913 letter to Bureau lens was also impacted by if, or how many, pul- of Lighthouses Com- leys were employed in the rigging of the cable missioner George between the clockwork and the weights them- Putnam, Horatio B. selves. Sizable adjustments to the rotational Bowerman, the chief speed of the lens could be made by adding constructing engineer or subtracting various poundage of weights. then engaged in re- Commonly these were iron disks, but an 1894 working the weight letter to the district engineer from the keeper of system at Cape Hat- the Heceta Head, Oregon, Lighthouse noted teras wrote in regard that “about 50 pounds of chain has been added to the existent sys- to the revolving apparatus.” Finer adjustment tem: of the speed was by means of the “vanes” that The keeper tells Post 1913 rail system for guiding weights – no longer in use – Cape were a part of the clockwork regulator. Tem- me that it has borne Hatteras, North Carolina. Photo by Jami Lanier, .

The Keeper’s Log—Winter 2016 37 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 How Frequently Were The winding interval was also affected by the simple fact of how tall the tower was and Weights Wound? how far up and down that tower the weights ust as lighthouse folklore often had the keeper were actually allowed to travel. Jlugging unbelievably unnecessary amounts At the Cape Meares, Oregon, Lighthouse, of fuel up the stairs to the lantern, so too do it is only about 10 feet from the underside of today’s tour guides like to talk of the unending the pedestal deck to the floor below. Subtract labor by the keeper to wind those heavy weights the length of the weight unit itself and there back up to the top of the lighthouse. were only a few feet that the weights could The amount of physical labor required un- travel before they had to be wound. It would doubtedly varied with the particulars of the appear the winding interval was quite short equipment at any given site. Francis Jacker, here, yet the inspection report from 1910 1880s keeper at the Raspberry Island, Wis- notes a 2½ hour wind time. The weights here consin, Lighthouse, noted in his log, “Proper obviously descended very slowly. surveillance of the revolving apparatus dur- The tower at Cape Hatteras, North Caro- ing the long nights of the fall, when frequent lina, is the tallest in the U.S., thus one would windings are required, is exhausting.” think that once wound up the weights here In most instances, however, the job of could descend a long way over a long period winding even the heaviest weights was made of time. This is not the case. The weights here relatively easy by the mechanical advantage could descend all the way to the bottom. of the gearing in the clockwork. It was sel- They were not wound all the way to the top, dom a hard job and in most cases did not Open top of weight tube, Cap Madeliene, however, but rather only a short distance up have to be done that often. Quebec, Canada. Photo by the author. from the bottom. (See more information on The inspection reports that were reviewed this system in a section below.) frequently but not consistently noted a time Hansen had to wind the weights 80 feet up As originally installed, the weights at the interval between windings. What is unclear from the bottom of the tower. In practice he Heceta Head, Oregon, Lighthouse could is whether these times are how long it would likely didn’t let them get down so far. only descend a portion of the tower height. have taken for the weights at that individual A few reports include comments that The keeper complained in an 1894 letter to lighthouse to go from the uppermost po- make it clear that at some sites the weights the district engineer that “if the tower had sition after winding to the lowest possible were arbitrarily rewound sooner than abso- been arranged to permit the weights to come point in the tower, or whether they repre- lutely necessary. The 1913 report for North down to the ground floor, it would run two sent the usually shorter interval at which the Manitou, Michigan, notes that the wind or three hours without rewinding.” At some keeper arbitrarily decided to rewind them. time was “5 hours, machinery wound every later date, holes were cut into the landings Even with the efficient gearing, it made 2.5 hrs.” A similar note in the 1912 report to allow a greater drop distance. sense to wind them a shorter distance more for Sanibel, Florida, notes that the weights The inspection reports from the- Na often than “all the way up” less often. “can run for 12 hours” but that they were tional Archives show a range of wind times, The opening paragraph says that Keeper usually rewound every four. the shortest being 1.2 hours and the longest

LIGHTHOUSE YEAR ORDER POUNDS ROTATION DROP TUBE FT. WIND BB, TIME TIME HR. CW, MF Barnegat, New Jersey 1910 1 4 M only part used 2 CW Cape Cod, Massachusetts 1927 1 200 20 s 2.25 MF Cape Fear, North Carolina 1918 1 500 4 M 2 CW Cape Hatteras, North Carolina 1927 1 150 2' 24" 130 15 CW Cape May, New Jersey 1907 1 8 M 8 CW Cape Meares, Oregon 1910 1 4 M 10 2.5 CW Cape Mendocino, California 1907 1 8 M 3 CW Carysfort Reef, Florida 1912 1 5.33 M 11 3.5 BB Destruction Island, Washington 1910 1 4 M no tube 64' drop 5 CW

Partial table shows, for a sampling of first-order lenses, the reporting year; weight; rotation time; length of the drop tube, if any; hours between winding; and whether the mechanism is ball bearings (BB), chariot wheels (CW), or mercury float (MW). The full table can be found on the U.S. Lighthouse Society’s website at https://uslhs.org/history/lens-rotation/weights-and-measurements.

38 Clockworks—Winter 2016 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 The heading for the- in top of the tower, snapping off eighteen of the spection item regarding steps before crashing to the floor of the light- the weightways was “Drop house. Fortunately no one was climbing the Tube Ft.” The fact that the stairs at the time.” column is so headed leads In an incident at Jupiter Inlet, Florida, the me to believe that an en- weights were once wound too far up, causing closed “tube” of some kind them to break free and plummet downward, was the normal situation damaging five of the stairs as they fell. rather than an exception. Keepers likely kept a close eye on the Many of the entries sim- condition of the cables and replaced them as ply give a number (130, 45, necessary, but a fascinating glimpse into this 54, 80 as examples), which I is provided by the log books from the Point believe indicates the length Sur, California, Lighthouse. The log notes the of an actual enclosure for April 1875 receipt from the lighthouse inspec- the weights. Several entries tor of “one coil of rope for our revolving light.” (all text is quoted verbatim) The entry for September 5 of that year notes, state such things as “full “Put new cord on clock work.” Between that len. of tower” or “stairwell” date and September 1895, a span of 20 years, or “whole height,” which the log documents that a new cord was in- Weight system plan, Fire Island, New York. Plan courtesy of I also interpret to refer to stalled no less than 17 times, sometimes only Dave Griese. an enclosure. Several say a month apart, at other times several years “none” or “no tube,” which between replacement. Most often the word being 15 hours. The majority of the noted I interpret to mean the weights here hung “cord” is used, but sometimes the word is intervals are between two and six hours, openly down the tower. “rope.” Only the December 1889 entry notes with the average of the 91 times noted in The earliest weights were suspended on the words “steel wire clock cord.” the reports I reviewed being 5.4 hours. Ten hemp or other fiber ropes, and wear and strain Types of Weightways of the 91 entries show a wind time of 10 probably caused their breakage far more often hours or more. Even on the longest winter than we know. Fiber ropes were replaced by his section notes different types of night, it would seem that not many keepers “wire rope” cables when this innovation came Tweightways as they once functioned exhausted themselves winding the weights. along in the later 1800s, and these likely part- and/or still survive at individual lighthouses The Weightway ed less often. A few of many documented in- across the country. I am appreciative of and stances of the cable breaking are noted below. indebted to staff at the various lighthouses eightway is the proper term for the An 1865 letter from the Third District noted. This is by no means a comprehensive Wpath by which the weights descended Engineer to the Light-House Board notes list, but represents a sampling. the towers. Whatever their appearance, damage caused by the “break of the clock weights all hung on the end of a cable (often cord” at the Fire Island Lighthouse. “The referred to as a “rope” or a “cord”) that was clockweight which runs within the iron wound around a drum within the clockwork newell [center column] is stopped by means located at the base of the pedestal that held of two iron bars running crossways through the lens. This cable then went through an the newell. In consequence of the parting opening in the pedestal deck, and gravity was of the clockcord the falling weight broke the the motive force that caused the weights to two bars and also knocked pieces out of the descend. The unwinding of the cable turned newell, which is a main support of the iron the drum on which it was wound, turning stairway.” In a creative bit of engineering, a gear that turned yet other gears, until the it was recommended that, in addition to mechanical advantage of the gearbox ulti- repairing the damage, a “spring sufficiently mately turned the lens itself. strong to counteract the force of the falling Sometimes the weights were enclosed weight” be installed on top of the iron bars within a tube, a box, or a channel in the wall, to prevent similar damage should the cord or sometimes they just hung down the center break again. There is no record of such a of the tower openly from the clockwork. spring being installed. The most accurate and informative infor- New England and lighthouse historian Ed- mation about the various weightways came ward Rowe Snow, in his book Famous Light- from staff at the individual lighthouses not- houses of New England, references an 1869 Weight pit, Keldsnor Lighthouse, Denmark. ed, but some information was gained from event at the Petit Manan, Maine, Lighthouse. Photo from the Egbert Koch collection, cour- the inspection reports that were reviewed. “The heavy weights of the clock fell from the tesy of Thomas Tag.

The Keeper’s Log—Winter 2016 39 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 Weights Hung Down the In other cases, some lighthouses had guide “eyes” built into each landing to keep the cable Center of the Tower straight down the center. Yaquina Head, Or- his arrangement, in which the weight egon, and St. Augustine, Florida, are examples Tassembly simply hung openly right down of towers with such guide eyes. These guide the center of the tower is the simplest and was eyes at stations such as Yaquina Head kept the quite common in towers short and tall. This cable straight but restricted the weight assem- was most easily accomplished in towers where bly itself from being raised beyond the lower- the winding staircase was attached to the out- most of the guides. Thus when “wound” these side wall of the tower, thus leaving a large open weights were only a relatively short distance space in which the weights could hang. up from the bottom of the tower; they could A 1911 inspection report for Hillsboro not be wound all the way to the top. Inlet, Florida, states that no “drop tube” ex- In an even more unusual variation of a isted but that the weights simply “drop thru guide device, an early letter regarding the center of columnar stairs.” Cape Hatteras tower noted that a black A 1910 report from the Thunder Bay walnut board with a hole in it was used to Island, Michigan, Lighthouse says similarly, “keep the rope centered.” “No drop tube, weights drop down center A common feature of the towers in of tower.” which the weights hung down the center These are only a few of the many reports was a well-like pit, approximately four feet that note “none” regarding a drop tube. Above: Weight pit, St. Augustine Lighthouse, in diameter and several feet deep. This was Florida. In several lights there are guiding devices in the center of the floor and usually sur- Below: Guide “eye” on landing for the weight to keep the cable hanging truly in the center rounded by a railing. These pits are usu- cable, centered over the weight pit at St. Au- of the tower. These are particularly impor- ally found where a first-order lens, which gustine. Photos by Kathleen McCormick. tant in the taller towers where a long cable required many pounds of weights to turn with heavy weights on the end could easily it, was in a very tall tower. If the cable on begin to act like a pendulum. which the weights hung should break, the In some instances (Cape Hatteras, for weights could plummet a long way down example, as its weight system was changed and potentially do great damage to people in 1913), there were two rails that extended or objects in the way or to the tower floor down the center on which the weight as- itself. This pit, which was filled with sand, sembly traveled. provided a “cushion” onto which the falling In some cases other cables were used to weights and cable could land. guide the weights. Such was the case at the Ponce Inlet is one example of a tower Detroit River Lighthouse. A 1909 inspec- with such an enclosed pit on the bottom tion report noted “weights are held in place floor into which the weights would descend. in. The photo of the Keldsnor Lighthouse by two guide lines secured top and bottom The tower at Yaquina Head, Oregon, once in Denmark that accompanies this article by eye bolt.” had such a weight pit, but it has been filled shows the pit filled with sand. Pigeon Point, California; St. Augustine, Florida; and Cape Hatteras, North Carolina, are other light- houses that have such a pit. Shorter towers generally did not have a well pit; if the cable broke, the weights would simply hit the floor below. Undocumented lore at the Cape Meares, Oregon, Lighthouse, where the distance the weights could fall is less than 10 feet, tell of a crack in the stone floor caused by falling weights. Today’s guardians of that lighthouse deny such a crack, but it had to have happened someplace. Off Center but Still Hanging Openly variation of the “down the center” system Asometimes found the weights descending not exactly in the center of the tower, but Hole in stairs for weights, Jupiter Inlet, Forida. Photo by Josh Liller. slightly offset. An example of this can be found 40 Clockworks—Winter 2016 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 would have contributed many pounds. In 1913 the system was converted to weights guided by L-shaped rails and reconfigured so that the weights descended over a much greater distance in the tower. Centered but Inside the Steel Column That Supports the Circular Staircase

n towers where the stairs were attached Drop tube, Point Sur, California. Photo by Ito a central column, this column often Carol O’Neil. served as the weightway as well. This cen- tral steel column that supports the stairs is work weights. As noted earlier, I believe that called a “newel post” or simply a “newel.” some kind of such enclosure was the norm, The Burnt Island, Maine, Lighthouse is one although exactly where in the tower such such example. An opening at the bottom of tube was located varied greatly. These tubes the stair column allowed inspection of the were sometimes not in the exact center of the weights as well as retrieval should the cable tower but were offset or, perhaps, located just break. The Fire Island, New York, Light- inside the exterior wall. house had another example of this type of The drop tube at the Amelia Island, weightway, as did that at Sanibel, Florida. Florida, Lighthouse, is described by staff as a “wooden box on the inside of the wall.” Some were built partially within the out- Off-center drop tube, Point Arena, California. side wall of the tower. Photo by Ty Moore. An inspection report for Fenwick Island, Delaware, describes drop tube as “ wrought at the Pensacola, Florida, Lighthouse. Holes iron, about 8" in diameter, and projects cut in the pedestal deck and in one of the from wall of tower about ½ its diameter.” decks below allowed the weights to descend Spectacle Reef has a square such tube approximately 38 feet before they would have built against the outer wall. settled onto the next deck level below. The tube in the Point Reyes, California, A most unusual variation on this was Lighthouse is located about halfway be- once in use at the Cape Hatteras Light- tween the outside wall and the center of the house, well known as the tallest lighthouse tower. The weight cable here took an un- in the U.S. The distance from the floor on usual route—passing beneath the floor on which the clockwork, and the cable drum, which the clockwork sat, then coming up was located all the way to the sand pit at the into the watch room to a pulley above, and ground level is approximately 160 feet. That then entering the tube. number of feet of the typical wire rope cable Such tubes can also be found at Split in use in the early days at Hatteras would Central stair column that also serves as Rock, Minnesota; Rock of Ages, Michigan; have required a cable drum much larger weight tube, Burnt Island, Maine. U.S. Coast Point Sur, California; Elbow Reef, Bahamas; than could have been incorporated into the Guard photo. and Devils Island and Raspberry Island, clockwork there, so a very unique system Wisconsin, to name a few others. was created. Only about 50 feet of cable A classic example of a dedicated drop tube was actually wound onto the cable drum. Inside a Dedicated Tube is still in use at the Split Rock, Minnesota, Below the cable was an iron rod composed or Channel Lighthouse; the huge bi-valve lens here is still of 10 14-foot-long smaller rods that were being turned by the weights and clockwork. linked together—in effect a 140-foot-long nother type of weightway was an enclosed The tube is separate and distinct from the solid rod—on the end of which a series of Ametal or wooden column, referred to as spiral stairs, which are against and connected the common stackable weight discs were in- a “drop tube,” that went straight down the to the outside wall of the tower. It actually stalled. The rod system, of course, could not center of the tower from the pedestal deck descends several feet below floor level and has be wound up onto the drum, only that ap- level to or even to below the lower floor. This a door at the bottom, accessed via a hatch in proximately 50 feet of cable unwound as the was often the case where the spiral stairs were the floor. The 180 pounds of weights in use weights descended and was rewound by the connected to the exterior wall of the tower today can be inspected or adjusted via this keepers. It is not known how much weight rather than to a central post, and this drop door or retrieved in the case of a cable break. was used in this system; the rods themselves tube served exclusively to contain the clock- A second door is several feet above the floor The Keeper’s Log—Winter 2016 41 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 scribed in the same 1848 letter from Ste- was able to contact report no such exterior phen Pleasonton referenced earlier in regard tubes present, but whether or not such a tube to the weights at the Cape San Blas Light- once existed is no longer recalled. Many of house. The weightway which was to be con- these present-day caretakers suspect that the structed there was described as a “box ten weights in any that had a rotating lens must inches square in the center of the staircase.” simply have hung down the center, in spite of In an External Tube— the relatively tight quarters these central col- umns provided. The 1912 inspection report Skeletal Towers for the Sanibel Island, Florida, Lighthouse, hese “erector set” style lighthouses notes that the drop tube is a “newel post, 65 ft.” Tare characterized by a relatively small This is the center post to which the stairs are diameter central cylinder, which houses the attached; it served also to house the weights, so staircase, surrounded by a grid of iron/steel this was indeed the situation there. Plans of the , Califor- beams and rods, all surmounted by the lan- I suspect that in at least some other instances nia, show (above) location of column (drop tube) for weights and (below) the cable rout- tern assembly which itself looks little dif- there once was an external weight tube but that ing. National Park Service photos courtesy of ferent than the one on other lighthouses. The it was removed as a no-longer-needed and un- Carola Derooy. diameter of the central column provided little necessary maintenance headache when an elec- room for anything other than the stairway. tric motor relegated the weights to the scrap The tower at Whitefish Point, Michigan, heap, but documentation of this is lacking. includes a separate drop tube, external to and Inside the walls parallel to the central stair column, in which the weights that once turned a lens there de- he stately tall brick towers that are found scended. The cable passed out through the Tin many locations along the shores of wall of the watchroom, where the clockwork Lake Michigan and Lake Superior are often unit was located, and entered the tube at the referred to as “Poe Towers,” as their elegant gallery deck level. This system is no longer in design is attributed at least in part to early use, but the tube remains. A July 1909 let- lighthouse District Engineer Orlando Met- ter from lighthouse engineer Ralph Tinkham calfe Poe. These lighthouses are double- describes this drop tube at Whitefish as “60' walled, with the outer tapering cone of brick 5" long, is cylindrical sheet iron shaft 12" x surrounding an inner perfectly cylindrical 15" leading down outside of stair cylinder wall. The inner and outer walls are con- from lantern deck to ground.” nected, not separately free-standing, but this Photos of the skeletal towers at Detour construction results in an air space between level. This door was replaced with a glass Point, Michigan, and Manitou Island, the two which itself tapers from several feet window to allow visitors to see the weights Michigan, show that such separate exterior at the bottom to several inches at the top. as they descend. The window also alerts staff tubes were at these lights at one time also, It is in this space between the walls that the that it will be another 45 minutes before the but I have turned up no evidence of others. weights in those towers that had a rotating weights bottom out and the lens will stop Staff at several other skeletal towers whom I lens were hung, and the construction of them turning. The weights are intentionally al- provided for that potentiality. Towers at Au lowed to settle on the bottom of the tube Sable, Michigan; Big Sable Point, Michigan; when staff wants to stop the lens rotation. Little Sable Point, Michigan; Outer Island, Typically the cable, with the weights hung Wisconsin; and Cana Island, Wisconsin, are on the end, would pass through an opening a few of the many examples of these. in the pedestal deck right below the clock- Special access doors into the between- work and enter the tube directly, but there the-wall space were installed in most of these were variations. In one variation, the cable towers when they were built, even though on which the weights at the Fire Island, New some of them never displayed a rotating lens. York, Lighthouse were suspended did not The upper door was located just below the go directly from the drum on the clockwork pedestal deck and had a vertical slot through into the central tube, but rather first went to which the cable entered the wall. The cable a pulley mounted on the outside wall and then went around a pulley inside that upper then back to the central column. As in many door and the weights hung down from there. things regarding lighthouses, one would A lower door, without the need for the slot, think there would be more uniformity, but was located some distance below. Should the the variety of creativity shown by the engi- cable break, the weights could be retrieved neers and builders is surprisingly extensive. External drop tube of a . Mani- for reinstallation via this lower door. An early example of a drop tube is de- tou Island, Michigan. U.S. Coast Guard photo. With the exception of overall height, 42 Clockworks—Winter 2016 Reprinted from the U. S. Lighthouse Society’s The Keeper’s Log – Winter, 2016 A unique additional feature of several of and Pensacola and Jupiter Inlet, Florida, are these towers is a horizontal slot in the wall, some examples. a few feet below the upper door. A wooden An unusual variation of the cable routing or metal “shelf” could be inserted into this once existed at the Jupiter Inlet Lighthouse. slot. During the day, or when work had to When first put into service, the weights of the be done on the clockwork, this shelf would clockwork did descend in an inner-wall space, be inserted and the weights would be low- but this channel ended at the third landing of ered on to it, thus taking the strain off the the tower, at the 75th step. In 1928 the sys- cable drum of the clockwork. tem was changed, and the weights were rehu- The lighthouse at Outer Island, Wiscon- ng just inside the inner wall. An opening for sin, exhibits a unique feature not found in them to pass through was cut into the 92nd any of the other similar towers—a third ac- step. The hole, now with a plexiglass cover, cess door into the weight channel. This one is still there, intriguing today’s visitors. In a is located just below the upper door where variation of “within-the-wall,” the relatively the cable entered the chamber and is signifi- short “schoolhouse” tower at White River, cantly longer than the upper one. When the Michigan, has the weightway built into an weights were wound to the uppermost posi- otherwise solid masonry wall. tion, they would have been hanging right Summary Upper doors into inter-wall space at (above) inside this door and easily accessible for ei- Cana Island, Wisconsin (photo by Lee Radzak) ther inspection or for adjustment. If this was lthough it has been decades since elec- and (below) Outer Island, Wisconsin (National the reason for this third larger door, I have trification led to the mechinization of Park Service photo by David Cooper.) A no explanation as to why it is not found in lighthouses, and soon spelled the death-knell any of the other similar towers as well. of the keepers themselves, many remnants Double-walled towers exist at many oth- of the weight-driven clockwork systems that er locations besides the Great Lakes. Mon- once turned the rotating lenses survive in tauk, New York; Cape May, New Jersey; lighthouses worldwide, and more and more of today’s caretakers of these icons of our mari- time history are working to save if not utilize these remnants. A precious few sites still use this technology, and they provide a valuable glimpse into “what was” for a century-plus. Among the many nightly tasks of Keeper Hansen at Crooked River, and all of his peers at lighthouses from coast to coast, “winding the weights” was but one job, al- beit among the most critical. As is true in al- most every aspect of lighthouses, the variety and creativity of the engineers in designing weightways, through which up to hundreds of pounds of driving weights descended the towers, is astonishing. It is my hope that this article will bring these towers are in most other respects al- new interest to this almost forgotten part of most identical, yet there is variance as to the lighthouse history. distance between the upper and lower doors, and thus the distance the weights could de- Particular thanks for information and/or scend before they would have to be wound photographs to Deb Bender, Michelle Buckley, back up. The upper and lower doors at the Chris Case, George Collins, David Cooper, South Manitou Island, Michigan, Light- Dave Griese, John Havel, Matt Kuehne, Jami house are about 40 feet apart. In the tower Lanier, Josh Liller, Bruce Lynn, Eric Mar- at Au Sable Point, Michigan, the two doors tin, Kathleen McCormick, Ty Moore, Beverly are about 61 feet apart. In yet another very Mount-Douds, Carol O’Neil, Henry Osmers, similar tower at Little Sable Point, Michi- Terry Pepper, Lee Radzak, Thomas Tag, Matt gan, the two doors are 78 feet apart. These Upper door opening into inner-wall space Varnum, and Jim Woodward. distances represent the maximum distance (door removed), showing cable going around the weights could have descended, and I the pulley inside. Little Sable Point, Michigan. have no explanation for the differences. Photo by Matt Varnum, SPLKA. The Keeper’s Log—Winter 2016 43